getMj: An Estimator for error-covariance matrix.

In DylanSpicker/rcalibration: Implements Generalized Regression Calibration (and Related) Estimators for Measurement Error Correction.

Description

The generalized estimator works on error models with covariance of the form cov(X) + Mj; this function estimates the residual Mj term.

Usage

getMj(W, enforce.psd = FALSE, ...)

Arguments

W	A list of length 'k' containing matrices of error-prone proxy measurements of the covariate. Matrices should all be n (observations) x p (dimension of covariates).
enforce.psd	Should the returned matrix be projected to be positive semi-definite. Can be a list (of the same length as W) or a boolean. If TRUE, the corresponding Mj will be computed as the eigen-decomposition of Mj, with eigenvalues cast to 0. Defaults to FALSE.

Value

A list M_j, the error-covariance structure matrix.

See Also

[rcalibration::getOptimalWeights()] which calls this function to compute weights [rcalibration::generalizedRC()] which uses this if non-optimal weights are selected

Examples

getMj(W)

DylanSpicker/rcalibration documentation built on March 8, 2020, 10:38 a.m.